Well to find the answer Richard visits Professor David Hopkinson a geneticist from University College London, who’s been analysing hundreds of faces over the past two years…

Richard: I’m assuming this is some sort of scanner, so what are you doing with it, and, why?

David Hopkinson: This is a 3D face scanner and what we’re trying to do is to capture something like thirty thousand data points, which would allow us to construct this person’s face, in a 3D image on a computer screen. What we’re trying to work out, is the relationship between the shape of her face, and her genetic constitution.

Richard: Why have you got Shelly’s son and mother here?

David Hopkinson: To find out how much influence our genes have on our appearance and see which facial features are inherited factors.

Richard: What use is this going to be?

David Hopkinson: Well, it’s useful in the medical sense, because it gives us information about normal features development, which might have some impact on studies of abnormalities of development. But also there is a forensic use, if for example, at a scene of crime, we took a DNA sample from a blood stain or other biological sample to analyse we could make some prediction about the appearance of that individual.

Richard: You mean even from a blood stain you could tell what someone looks like?

David Hopkinson: Yes in principle.

If you would like to find out more about the use and meaning of visual images then have a look at course D850 The Image and Visual Culture
But what about looking at our faces in a mathematical way, how can our appearance have any relevance to numbers, Richard meets Dr Chris Solomon at Kent University…

Dr Chris Solomon, Kent University

Richard: So tell me Chris, why does the way we look all boil down to a mathematical calculation?

Chris Solomon: The basic idea Richard is that we start with an average face. The face you see here is in fact the average of a sample of about five hundred faces. Essentially when you have a digital image, the grey scale is mathematically quantified. So for example, dark areas in the image tend to have low values. You then average the number lying in each of these groups.

Richard: This image looks too perfect to be average to me.

Chris Solomon: The subjective impression of the attractiveness of a face is in fact related to how close it is to the average face. In other words the closer a face is to the average face, it’s actually perceived as being more attractive.

Richard: Is it possible then, that there could be someone else who looks exactly like me?

Chris Solomon: I suppose from the mathematical point of view it’s possible, although I have to say that statistically speaking it’s unlikely.

Richard: What uses can this system have for us?

Chris Solomon: If it’s possible to specify somebody’s face by a combination of these faces, then all we really need to know is the set of numbers, which can tell me how much of each face I need to produce, we can then use this for numerous purposes. Such as a credit card, where there is just enough space to store the facially encoded image on the magnetic stripe. So for example when you present your card to the cashier to pay for something, as soon as they swipe the card through a magnetic stripe they can match the image, which is produced by the card, to the image of you, the holder of the card.

Richard: So you’re saying Chris that a face can be boiled down to a mathematical equation.

Chris Solomon: That’s absolutely correct yes.

Richard: I’m not sure that I’d want my face imprinted on all my credit cards even although I understand the security implications, but I do think I’m beginning to understand, how it is that I can recall or remember a face, even though I don’t remember the name.

    next > Page 1 of 2

Content last updated: 15/03/2005

Comments